pickle error in load_tf_model

[jdc891010]

/Encountering a "ValueError: unsupported pickle protocol: 5"

When trying to run the example code:

import pyoxynet

# Load the TFL model
tfl_model = pyoxynet.load_tf_model()

# Make inference on a random input
pyoxynet.test_tfl_model(tfl_model)

# Plot the inference on a test dataset
pyoxynet.test_pyoxynet()

forums mention that it has to do with the python env in which pickle was created, I am running it in 3.7, and from the DockerFile it seems that you prefer 3.8?

A possible easy fix, that I tested, is using the "import pickle5 as pickle" for py < 3.8

[andreazignoli]

Cannot replicate the issue locally. I add a sentence in the README specifying this was developed under 3.8.

[andreazignoli]

Hello @jdc891010, any luck in installing locally a 3.8 version perhaps with pyenv?